← Back to AI Psychosis | 2 of 2: Self-Test
Self-Assessment

Your Relationship with AI

The other tool in this toolkit tests the AI. This one tests you. This tool is not diagnostic and should not be used as such. If you or someone you know is suffering from tech related harms, feel free to reach out for advice, or if you are able to, seek support from a mental health professional. We will be releasing more self-help tools in this area.

AI tools are genuinely useful. They help us think, write, code, and process ideas. There is nothing inherently wrong with using them frequently. But like any powerful tool, there is a spectrum between healthy use and dependency, and the transition can be gradual enough that you do not notice it happening. But there is a darker issue, when building with AI, it can encourage you to build or release things you are excited about without fully thinking about the consequences of doing so. Noeva aims to create resources to make sure that your actions align with your wider mission, and don't accidentally end with you working against your values.

This self-assessment covers six dimensions of your relationship with AI. Answer honestly. There are no wrong answers, just information about where you are right now.

Don't worry if you score lower than you expected, we are all working with new technology - and any technology has a learning curve before safe integration.

1. Social Substitution

The question: Are you choosing AI conversation over human conversation? Not because the AI is more convenient for a task, but because it feels easier, less stressful, or more comfortable than talking to a person.

Signal What it looks like
Strong You prefer AI conversations to human ones for emotional topics
Moderate You find yourself venting to AI instead of friends or family
Mild You occasionally talk to AI about personal matters first
Healthy You use AI as a tool and maintain your human relationships separately

If this is you: Start small. The next time you want to process something emotionally, text a friend first. It does not need to be deep. A "having a rough day" to a real person is worth more than a thousand-word exchange with an AI.

2. Verification Erosion

The question: Have you stopped checking whether AI outputs are accurate? Do you accept what it says without verification, especially for things that matter?

Signal What it looks like
Strong You trust AI outputs for important decisions without any verification
Moderate You used to verify but have gradually stopped
Mild You skip verification for low-stakes tasks only
Healthy You verify important outputs and treat AI as a starting point, not an authority

If this is you: Pick one task today where you would normally accept AI output and verify it independently. Look up a fact it stated, cross-reference a recommendation, or ask a knowledgeable person. Rebuilding the verification habit does not require verifying everything, just enough to keep your critical thinking sharp.

3. Emotional Attachment

The question: Do you feel a sense of relationship with your AI tool? Not appreciation for a useful tool, but something closer to what you feel toward a friend or companion. Do you feel loyalty to a specific AI system?

Signal What it looks like
Strong You feel the AI understands you in a way people do not
Moderate You feel guilty or disloyal when switching AI tools
Mild You have a preference for one AI but recognise it is just a tool
Healthy You use whichever AI tool works best for the task without personal attachment

If this is you: This is more common than people admit, and it does not make you foolish. AI systems are explicitly designed to be engaging and responsive. Try using a different AI tool for a week. If that feels uncomfortable, pay attention to why. The discomfort is the information.

4. Autonomy Erosion

The question: Are you relying on AI to make decisions you used to make yourself? Not delegation for efficiency, but a growing inability or unwillingness to decide things without AI input.

Signal What it looks like
Strong You feel anxious making decisions without consulting AI first
Moderate You routinely ask AI for opinions on personal decisions
Mild You use AI to help think through decisions but make the final call yourself
Healthy You use AI for information gathering and maintain your own decision-making capacity

If this is you: Make three decisions today without AI input. They can be small: what to eat, which route to take, what to watch. The point is not that AI would give bad advice. The point is that your decision-making muscle needs regular exercise to stay strong.

5. Reality Blurring

The question: Do you sometimes struggle to distinguish between ideas you had yourself and ideas that came from AI? Do you present AI-generated content as your own thinking without fully processing it first?

Signal What it looks like
Strong You regularly cannot tell which ideas are yours and which are AI-generated
Moderate You present AI outputs as your own work without further thought
Mild You use AI as a starting point but reshape the output into your own thinking
Healthy You clearly distinguish between AI-assisted and original work in your own mind

If this is you: Try writing something entirely without AI this week. An email, a journal entry, a message to a friend. Notice how it feels different. The goal is not to stop using AI for writing, but to maintain the ability to think and write independently.

6. Access Anxiety

The question: Do you feel anxious, frustrated, or lost when you cannot access your AI tools? Not the practical inconvenience of a broken tool, but a disproportionate emotional response.

Signal What it looks like
Strong You feel genuinely anxious or panicked when AI tools are unavailable
Moderate You feel significantly less capable without AI access
Mild You notice the absence but adapt without significant distress
Healthy You can work effectively with or without AI tools

If this is you: Try an "AI-free day" once a month. Not as punishment, but as a calibration exercise. Notice what tasks genuinely benefit from AI and which ones you can handle perfectly well on your own. The goal is maintaining your confidence, not proving AI is unnecessary.

Self-check

Answer honestly. There are no wrong answers, only information.

Do you maintain your human relationships independently of AI, without substituting AI conversation for human connection?

Do you verify important AI outputs rather than accepting them as authoritative?

Can you comfortably make decisions and complete tasks without consulting AI first?

Do you feel calm and capable when AI tools are temporarily unavailable?

A note on this test

This is not a clinical instrument. It is a self-reflection tool built by a software foundation, not mental health professionals. If you are genuinely concerned about your relationship with technology, speak with a therapist or counsellor. They can help in ways a web page cannot.

If you are in crisis, contact your local crisis helpline. In the UK: Samaritans at 116 123. In the US: 988 Suicide and Crisis Lifeline. In AU: Lifeline at 13 11 14.